Abstract Mitochondrial dysfunction has been proposed as a major factor in insulin resistance, aging, and metabolic diseases. 13C NMR in vivo has been the main method to assess mitochondrial fluxes like the TCA cycle and anaplerosis. NMR measures 13C flow from labeled substrates like [3-13C]lactate into the amino acids aspartate and glutamate, with the rationale that via 13C exchange with the TCA intermediate ?-ketoglutarate, glutamate is a ?trap? for 13C mixing with TCA cycle intermediates. Because NMR in vivo requires major technical expertise, methods exist to measure plasma glucose labeling from precursors that enter hepatic metabolism and, from steady-state C labeling, estimate VTCA, particularly using C-propionate. Although in principle these methods 13 13 should agree with tissue measurements, large discrepancies have been observed in several rates, including VTCA. The divergence is the subject of several recent commentaries, letters, and symposia but lacks a clear resolution. Resolving the controversy is key to understand the role of mitochondria in the pathogenesis and treatment of hepatic insulin resistance, nonalcoholic steatohepatitis, and type 2 diabetes. A solution to the controversy is to measure 13C positional labeling of TCA cycle intermediates. NMR in vivo and steady-state plasma glucose methods yield indirect measures of mitochondrial metabolism and depend on some incompletely tested assumptions about relationships with cytosolic glutamate and aspartate. We recently published the Mass Isotopomeric Multi Ordinate Spectral Analysis (MIMOSA) platform for comprehensive, stepwise, integrated analysis of intracellular metabolism (see Alves et al., Cell Metabolism, 2015). The ?mass isotopomer? aspect of MIMOSA uses MS/MS-based ion fragmentation analysis of stable- isotope-labeled metabolites to identify carbon-specific label positions. The ?multi-ordinate? aspect is a major innovation that allows direct assessment of label flow along intersecting pathways, including mitochondrial intermediates that are inaccessible by positional NMR due to sensitivity limitations. We used MIMOSA in a cell model and found that previous measures of anaplerosis by steady-state glutamate labeling were up to 3x too high due to mitochondrial dilution pathways that could not otherwise be measured. We propose to apply MIMOSA in an animal model in vivo to establish the ground truth for hepatic VTCA and other key fluxes (Aim 1). We will use the information to test the accuracy of present methods used in vivo for human and rodent studies (Aim2) and develop improved measurement methods. Aim 3 will assess plasma labeling patterns resulting from tissue-specific metabolism that can impact the interpretation of tissue data. Our preliminary data identify a lactate-glycerol shunt in adipose that may have pathologic effects in addition to confounding flux measurements in vivo. Consequently, targeting this pathway may be a novel treatment for diabetes or other metabolic diseases. A major translational goal is to develop a cross-validated in vivo analytic platform using either MS or NMR either humans or rodents.